| | 5 | |
| | 6 | == WP4 innovations available through the AmbientREP == |
| | 7 | |
| | 8 | * Self calibrated localization: Real-time localization of mobile nodes in a dynamic environment with self calibration. Benefits are: |
| | 9 | * adapting to the environment and changing conditions on-the-fly without taking extra measurements |
| | 10 | * automatically calibrate nuisance parameters (path-loss exponent and reference RSS) from localization measurements, instead of a pre-calibration phase which is required in most other RSS-based methods |
| | 11 | * Activity monitoring (as part of activity recognition and monitoring): The activity level (or IMA value) is calculated on inertial sensor nodes worn by persons, to indicate the amount of movement of that person. Doing this calculation on the nodes saves bandwidth since the raw inertial sensor data would consume much more bandwidth. Additionally, applications do not have to deal with raw sensor data which is often sensor-specific. Other benefits of activity recognition and monitoring are: |
| | 12 | * scales well with number of sensors/actuators |
| | 13 | * high accuracy and low false alarm |
| | 14 | * distributed and real-time |
| | 15 | * low memory and processing power are required |
| | 16 | * Outlier detection (as part of outlier and event detection): identifying incorrect measurements within the sensor network and correcting them either by removing them or use them for further analysis helps in saving energy and ensuring data quality. Benefits of outlier and event detection are: |
| | 17 | * robustness against various failures, i.e., sensor, node, or communication link failure |
| | 18 | * adaptive and able to learn |
| | 19 | * low memory and processing power are required |
| | 20 | * high accuracy and low false alarm |
| | 21 | * distributed and real-time |
